Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Psychology



First Advisor

Wayne F. Velicer


Five computer programs that perform Box and Jenkins (1970) ARIMA model interrupted time series analysis, i.e., TSX, GENTS, SAS, BMDP, and ITSE, were evaluated both qualitatively and quantitatively. The qualitative evaluation reviewed their documentation and computational features. The quantitative evaluation was performed by comparing analysis results (estimates of the minimum residual error variance and the intervention parameters) for 10 replications each of 44 simulated time series. The 44 series include 11 different ARIMA models and degrees of dependency; by two forms of intervention, i.e., 1) level (L) and change in level (DL) alone, and 2) level (L), change in level (DL), slope (S), and change in slope (DS); and by two series lengths (40 and 100 points). Major findings are 1) differencing and some autoregressive model solutions are highly inaccurate even with the correct model indentification, 2) GENTS, which does not require that the model be identified, performed comparably (i.e., it provided equivalent solutions, both good and bad) to the "true model" solutions, and 3) model identification appears still to be necessary. Recommendations are made for current analyses, improvements to computer programs, and future research.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.